The contributions of generative artificial intelligence (AI) and social media to the climate crisis are often underestimated. To date, much of the focus has been on direct emissions associated with the life cycle of tech products. In this forum article, we argue that this narrow focus misses the adverse and indirect impacts of generative AI and social media on the climate. We outline some of the indirect ways in which generative AI and social media undermine the optimism, focus, creativity, and veracity required to address the climate crisis. Our aim is twofold. First, we seek to balance the tide of optimism about the role of digitalization in addressing the climate crisis by offering a skeptic’s perspective. Second, we outline a new research agenda that moves beyond counting directly attributable carbon emissions and proposes a more comprehensive accounting of the indirect ways in which social media and generative AI adversely impact the sociopolitical conditions required to address the climate crisis.

The COVID-19 pandemic is the largest public health crisis in recent history. Many states have taken unprecedented action in responding to the pandemic by restricting international and domestic travel, limiting economic activity, and passing massive social welfare bills. This begs the question, why have states taken extreme measures for COVID-19 but not the climate crisis? By comparing state responses to COVID-19 with those to the climate crisis, we identify the crisis characteristics that drive quick and far-reaching reactions to some global crises but not others. We inductively develop a conceptual framework that identifies eight crisis characteristics with observable variation between COVID-19 and climate change. This framework draws attention to under-considered areas of variance, such as the perceived differences in the universality of impacts, the legibility of policy responses, and the different sites of expertise for both crises. We use this structured comparison to identify areas of leverage for obtaining quicker and broader climate action.

Nonstate and subnational climate governance activities are proliferating. Alongside them are databases and registries that attempt to calculate their contributions to global decarbonization. We label these registries “orchestration platforms” because they both aggregate disparate initiatives and attempt to steer them toward overarching objectives such as improved transparency, accountability, and effectiveness. While well-intentioned, many orchestration platforms adopt a narrow conception of “value” as either quantifiable greenhouse gas (GHG) reductions or relevant outputs. We offer a more comprehensive approach to valuing nonstate and subnational climate governance that is rooted in recognizing the potential for initiatives to become far-reaching (i.e., achieve scale) and durable (i.e., become entrenched). We illustrate the comparative advantage of our approach with reference to a particular case of nonstate governance: The Carbon Trust’s attempt to create product carbon footprints. By tracing the direct and indirect impacts of product carbon footprinting, we show that initial failures to generate quantifiable GHG reductions or produce relevant outputs do not reflect the intervention’s broader impacts through scaling to other jurisdictions and entrenching business practices that contribute to decarbonization. Taking this broader view of “value” can help policy-makers better understand and gauge the contribution of nonstate and subnational climate governance to global decarbonization.